Pluralis Research
Research Engineer - Post-Training
San Francisco +1 · Remote · Full-time · Posted Aug 31, 2026
- Location
- San Francisco
Australia - Visa sponsorship
- Optional full visa sponsorship and relocation support to Australia or the US.
About the job
Develop end-to-end reinforcement-learning post-training algorithms and systems for decentralized models operating across consumer GPUs and Macs over the public internet.
What you'll do
- Build the RL training loop, including rollout ingestion, reward computation, policy updates, and distributing updated weights.
- Adapt RL algorithms to asynchronous, high-latency, partially trusted rollouts.
- Create evaluations and release the first decentralized post-trained models as public artifacts.
What they ask for
- Hands-on LLM RL post-training experience (RLHF, RLVR, or reasoning-focused RL), including systems work.
- Production-quality Python and PyTorch skills, including concurrency, failure handling, and profiling.
- Research ability demonstrated through relevant publications or detailed, defensible unpublished work.
- Belief in Protocol Learning as a viable approach to collective, trustless, and sovereign AI.
Nice to have
- Experience with slow networks or decentralized or federated setups; serving-engine internals such as vLLM or SGLang; reward modeling or verifiable-reward datasets; or P2P networking and NAT traversal.
- Experience at proprietary, open-weight, and open-source AI labs.
Tools
Python, PyTorch, vLLM, SGLang
AI focus
RL post-training for agentic use cases, including adapting algorithms and systems to decentralized, asynchronous training and inference.
Model-written summary · the original posting is authoritative